Becoming a data-driven organisation can help create a better shopping experience for your customers and drive growth through better decision making. Learn how to ensure your business is 'data-fit' for the future.
1. The future of data for retail
Jason Nathan
Global Head of Data
2. 22
The size and complexity of big data continues to amaze
MULTIPLYING
SOURCES
EXPONENTIAL
VOLUME
INCREASED
COMPLEXITY
& VELOCITY
Credit Cards
Company
Systems
Internet Browsing /
Clickstream
Social
Networks
Mobile &
Smart
Devices
Applications
Global Production of Data
Zettabytes pa
Mobile & Online Browsing Data
Exabytes pa
+40
%
+37
%
“internet of things”
Complexity
Videos Reviews GeolocationImages Daily
Real
Time
Velocity
3. 3333
I am going to concentrate on three facets:
Privacy and
the future of
the Data
Marketplace
Organisational
Data
Capability and
what it
enables
Building high
value Data
Assets
5. 55
Abilitytocapturesuperiorvalue
Capability level
LEVEL 1
LEVEL 2
LEVEL 3
LEVEL 4
LEVEL 5
Advanced
Data
Capturing and
utilising data
beyond Core
management
(i.e. Customer,
digital and
competitor
data.
Joined Data
Data joined
using
Customer,
Product, Store
keys to allow
for deeper
analysis of
behaviours
and
performance.
Enriched Data
Data joined and
enriched
(segmentations,
scoring)
continuously;
enrichment used
for reporting,
analytics and
activation.
Data
Partnerships
Enriched data
used to drive a
self-sustaining
commercial
relationship
ecosystem.
Both enables
the
consumption
of new data
(i.e. social)
and
monetisation
of existing
organisational
data.
Essential Data
Core Sales,
Product and
Stores data
captured and
used for
reporting.
The history of data in retail is reflected in the
organisational capability of retailers
6. 66
A robust data strategy is the foundation for
unlocking the value of customer data
dunnhumby uses six dimensions to evaluate a
retailer’s fitness for the future of data
9. 9
lifestyle
what
who
where
life stage mission
services
out of
store
needs
online
day of
week
how
offline
Partnerships enable us to better understand and engage
throughout the shopping eco-system
remind
delight
selection
save time
curate
inform
flexibility
reward
loyalty
ask
comment
notice
consider
need
pick
paycheckout
obtain
use
share
options
surprise
rewards
new
products
recognize
save
money
reject
segment
search
pick
visit
time of
day
in-store
interests
time of
year
when
behavior
plan
rewards
Multichannel Shop Ecosystem v2.0
occasion
connect
surprise
community
discover shop buy reflect
appreciate
help
products
11. 1111
example: using clickstream data to understand how
customers shop a category online
11
Insights:
Over and under performance online
Breakdown of purchasing across key site areas
Breakdown of browsing across key site areas
Benchmarking versus total Groceries
purchasing behaviour
Actions:
Create a priority list of category opportunities
Create case for taxonomy change
Support development of annual media planning
12. 1212
example: optimising search results, based on customers’
real search terms
12
Insights:
Excel based tool
Product performance within the top category
search terms
Top search terms by product
Importance of search to your products
Actions:
Optimise product descriptions (Brandbank)
Use common misspellings to identify search
thesaurus entries
Prioritise acquisition of sponsored search terms
(where available) and inform SEO
Inform brand messaging based on customer
language
14. 14
Transparency through simple and clear
T’s and C’s
Level and type of activity – relevant offers
and content , right frequency, appropriate
channel
Security needs to be the top hygiene
factor – rigorously embedded across the
organisation
Security, transparency and activity
are key to customer trust
Transparency through clear messaging of
usage of their data and the value it delivers
Pricing is a
particularly
sensitive topic for
customers so
transparency is
critical to earning
trust
15. 15
A growing number of start-ups have identified the
opportunity
Connect, Learn & Earn:
“Unlock the power of your data, the world’s first personal data marketplace”. Connect social
networks, banking, fitness, google+ and earn based on flat rate.
“Rewarding you for your personal data. Discover what your data says about you and earn every
time it is used”. Connect social, mobile, todo and browsers - insights generated and rewards provided.
Handshake, facilitates direct conversation between companies and customers to gain insights and
negotiate rewards.
Connect for Convenience:
convenient management of loyalty programmes (120+), including points & rewards, and ability to use to
scan – logs in on customer’s behalf
“Life management platform” – connect social, comms, memberships, browser to make more convenient
and co-ordinated. Get insights e.g. “life timeline”, and in future option to sell data.
Input for Security & Control:
Not-for-profit “digital passport” to manually capture information, focused on convenience, security and
control.
+more in beta testing/pilot phase
Today I’m going to talk about “What we see as the world of data in retail”
Our role at dunnhumby is to help our retail partners realise the value of their data to improve their decision making and optimise their ability to communicate with and serve their customers in a better way.
We help them organise and enrich their data asset to deliver those services.
The increase in the sheer volume of data being generated is well documented and serves to highlight the importance of using data to drive better decision making.
If companies want to become data-driven organisations, they need to decide what data they want to use from their own data streams and what data they want to source from third parties.
The maturity of data in retail has increased dramatically over the past few years, with some retailers now changing organisational structure to cope with their data asset.
However, many organisations still operate with a similar mindset and organisational structure to how they might have done so 10 years ago. There is growing maturity in how they think of data as an asset for their business, but the talent base and organisational design that goes with that has lagged behind.
We use this model to give clients a method to evaluate and understand where they are on the data journey.
It’s not always a fast and easy road. Many companies are inching their way to becoming data-driven organisations, using data to inform decision making. However obvious the business benefits might seem, there is a choice – you can choose to opt out of the data universe. As an example, one major clothing retailer from the UK high street has decided they don’t want to use data other than for operational efficiency. They have no interest in using data to learn more about their customers – data is just used to run the core operations of their business.
Becoming a data-driven organisation requires careful investment, with a projected 7 figure spend needed to commit to progress up the maturity curve.
I’ll talk a little now about the levels in this data maturity curve:
Essential data – any modern retailer should have this – sales data, staff data, basic supply chain data.
Advanced data – Retailers at this level typically capture a whole range of data which has latent value, but are not really using it or aware of how to best combine it. Often it resides in an operational silo. We help to identify the value, stewardship, technical implementation and opportunities combining these datasets can bring about.
Joined data & Enriched data – This is where some retailers start to join the data meaningfully. It can be joined around customers, around stores, around time dimensions, around products, but requires some effort to put this data together for a specific end in mind. The principal of a “data lake” is set to replace the Enterprise Data Warehouse as a primary mechanic with which to do that. Great retailers add scores to products and customers for purposes of targeting, identifying propensity to lapse, etc. This is data enrichment.
Data Partnerships – Retailers at this level have a clear view on the partnerships they create with providers, such as social media platforms, and organisations that help them understand open data. We help connect them with partners who make up their eco-system of partnerships which help them understand their customers better.
These 6 principles are applicable to most businesses:
Robust data collection – Many companies don’t have the right terms and conditions in place at point of collection, but this needs to be at the heart of the design of your data strategy. Data collection should not be an afterthought in technical or legal terms.
Talent and team structure - People who are owners and stewards of data within an organisation, often report into IT and finance. They should be connected to Marketing or whomever owns the relationship with the customer.
Data Partnerships – Some companies have good partnerships, but these are often only leveraged for single purpose.
Revenue generation – Many companies are sitting on data that has value in and of itself and could be leveraged outside their organisation if the right terms are in place.
Data privacy – How clear are you about speaking to your customers about what is happening with their data?
Single view of the customer – Is your data organised in such a way that it enables the best opportunities for analysis and actionable insight?
This model helps retailers organise thinking around the data they are investing in, collecting, managing, and presenting.
The starting point for any conversation around data needs to be an understanding of what at a high level you want to achieve and then creating a clear model for how your customers interact with you on their shopping trip.
Key questions to ask: which of those touchpoints generate data that you own or can access through a 3rd party? Can you can enrich any of those touchpoints with additional data or technology which helps customers and generates data?
Here are some examples of companies who provide third party data which can be used to enhance your existing data asset.
There are data where it’s valid and useful to have some time after the event, for modelling.
But some data is really useful to have at a particular moment of time (e.g. geo-location data for reaching people at certain locations within store). Only focus on short loop data where you really need to.
Clickstream data helps to understand the navigational path through an ecommerce website. How does the customer move through the website?
Which customers always click on the promotional offer on the front page?
Which customers use the search function?
Which customers use the favourites function?
This data is very useful for allowing the retailer to better understand what the customer cares about most and helps deliver the right thing for them.
What do we mean by privacy and what should companies be doing about it?
Security should be a hygiene factor with investment and infrastructure needs in place to manage this.
T&Cs should be another key area of focus for the industry – A challenge that many will face: dealing with the issue of data use becoming so complex that it is difficult to explain in simple terms to customers?
The dichotomy of privacy: Customers expect fantastic experiences (e.g. Uber), but in order to do that, individuals need to allow sharing of their data for those experiences to manifest. Nervousness around sharing data is often brought about because consumers don’t feel they have transparency of usage.
An example of activity: Optimising pricing – consumers are sensitive about the notion of different people getting offered different prices for the same product (e.g. when buying flights online). Increased transparency from companies around how and why would help consumer understanding, but this is likely to remain a constraint for data and analytics for years to come.
Data portability is something that is inevitable for consumers. Increasingly consumers will want access to the data that is held about them, and legislation could turbo-charge this.
Plenty of organisations have now opened that data out through APIs that allow you as an authenticated user to see how your data is being used.
Organisations can put this data together to create meaningful value for other organisations and the individual who has empowered them to use and join that data up.
The whole principle of moving ownership of data much more directly into the hands of the individuals is going to be a major dynamic of the data marketplace in the next 5-10 years.